I agree that testability is vital, but I don't think time spent interpreting the data we have is wasted if it leads to a simpler model -- a paradigm shift that doesn't have any obvious testable conclusions can be the basis of models making new predictions (heliocentrism didn't lead to noteworthy testable predictions for decades), and the author admits the possible tests of quantum gravity they're advocating are "farfetched". Maybe the reason experts are advocating more theorizing than testing right now is we just don't have very promising theories at this point; maybe we have a theory that the sun is the center of the solar system, but it won't offer any better predictions than epicycles until we consider that the planets' orbits don't have to be round.